Article Modeling Human Cytomegalovirus-Induced Microcephaly in Human iPSC-Derived Brain Organoids Graphical Abstract Highlights d Human iPSC-derived brain organoids to model HCMV- induced brain malformation d ‘‘Clinical-like’’ HCMV strain impairs human brain organoid growth and structure d HCMV-infected brain organoids exhibit abnormal calcium signaling and neural network d HCMV-induced brain organoid abnormality can be prevented by neutralizing antibodies Authors Guoqiang Sun, Flavia Chiuppesi, Xianwei Chen, ..., Don J. Diamond, Felix Wussow, Yanhong Shi Correspondence [email protected] (F.W.), [email protected] (Y.S.) In Brief Sun et al. used human-iPSC-derived brain organoids to model the effects of HCMV infection on human brain development. They found that a ‘‘clinical- like’’ HCMV strain impairs brain organoid growth, structure, and neural network activity. Moreover, the abnormal brain organoid development caused by HCMV can be prevented by neutralizing antibodies. Sun et al., 2020, Cell Reports Medicine 1, 100002 April 21, 2020 ª 2020 The Authors. https://doi.org/10.1016/j.xcrm.2020.100002 ll
30
Embed
Modeling Human Cytomegalovirus-Induced Microcephaly in ......Human cytomegalovirus (HCMV) is a ubiquitous and highly adapted human pathogen that establishes lifelong latency in in-fected
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Article
Modeling Human Cytomeg
alovirus-InducedMicrocephaly in Human iPSC-Derived BrainOrganoids
Graphical Abstract
Highlights
d Human iPSC-derived brain organoids to model HCMV-
induced brain malformation
d ‘‘Clinical-like’’ HCMV strain impairs human brain organoid
growth and structure
d HCMV-infected brain organoids exhibit abnormal calcium
signaling and neural network
d HCMV-induced brain organoid abnormality can be prevented
by neutralizing antibodies
Sun et al., 2020, Cell Reports Medicine 1, 100002April 21, 2020 ª 2020 The Authors.https://doi.org/10.1016/j.xcrm.2020.100002
Modeling Human Cytomegalovirus-InducedMicrocephaly in Human iPSC-DerivedBrain OrganoidsGuoqiang Sun,1,5 Flavia Chiuppesi,2,5 Xianwei Chen,1,5 ChengWang,1 E Tian,1 Jenny Nguyen,2 Mindy Kha,2 Daniel Trinh,1
Hannah Zhang,1 Maria C. Marchetto,3 Hongjun Song,4 Guo-Li Ming,4 Fred H. Gage,3 Don J. Diamond,2 Felix Wussow,2,*and Yanhong Shi1,6,*1Division of Stem Cell Biology Research, Department of Developmental and Stem Cell Biology, Beckman Research Institute of City of Hope,
1500 E. Duarte Road, Duarte, CA 91010, USA2Department of Hematology, Beckman Research Institute of City of Hope, 1500 E. Duarte Road, Duarte, CA 91010, USA3Laboratory of Genetics, Salk Institute for Biological Studies, 10010 North Torrey Pines Road, San Diego, CA 92037, USA4Department of Neuroscience and Mahoney Institute for Neurosciences and Department of Cell and Developmental Biology, Institute for
Regenerative Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA5These authors contributed equally6Lead Contact
Although congenital infection by human cytomegalovirus (HCMV) is well recognized as a leading cause ofneurodevelopmental defects, HCMV neuropathogenesis remains poorly understood. A major challenge forinvestigating HCMV-induced abnormal brain development is the strict CMV species specificity, which pre-vents the use of animal models to directly study brain defects caused by HCMV. We show that infection ofhuman-induced pluripotent-stem-cell-derived brain organoids by a ‘‘clinical-like’’ HCMV strain results inreduced brain organoid growth, impaired formation of cortical layers, and abnormal calcium signaling andneural network activity. Moreover, we show that the impeded brain organoid development caused byHCMV can be prevented by neutralizing antibodies (NAbs) that recognize the HCMV pentamer complex.These results demonstrate in a three-dimensional cellular biosystem that HCMV can impair the developmentand function of the human brain and provide insights into the potential capacity of NAbs to mitigate brain de-fects resulted from HCMV infection.
INTRODUCTION
Human cytomegalovirus (HCMV) is a ubiquitous and highly
adapted human pathogen that establishes lifelong latency in in-
fected individuals. Although it is usually benign, HCMV infection
during pregnancy can result in viral transmission to the devel-
oping fetus, thereby causing irreparable birth defects in new-
borns.1,2 Congenital HCMV infection can occur following primary
maternal infection or result from non-primary maternal infection
due tomaternal re-infection or viral reactivation.3,4 It is estimated
that 0.5%–2% of all newborns worldwide are affected by
congenital HCMV infection.5,6 Around 10%–15% of congenitally
infected newborns are symptomatic at birth, and these infants
may suffer from neurodevelopmental deficits, including hearing
loss, intellectual disability, microcephaly, or cerebral palsy.2,7,8
In addition, congenitally infected newborns that are asymptom-
atic at birth may develop neurological illness later in life.2,7 In the
United States, children with long-term medical conditions are
estimated to be more frequently associated with congenital
HCMV infection than with other well-known childhood diseases,
CellThis is an open access article und
such as Down syndrome, fetal alcohol syndrome, and spina
bifida.9
Despite the recognition of congenital HCMV infection as a
leading cause of neurological defects in newborns, HCMV-
receptor (EGFR), and integrins (a3, a5, and b3), have been previ-
ously described to be either directly or indirectly involved in
HCMV host cell entry.39-41
To determine the importance of these cellular receptors for
brain organoid infection by TB40/E, we knocked down the
oid Growth by NAbs
P-labeled TB40/E in the presence of different concentrations of NAb 1B2 that
organoids in the presence of IgG control (2,000 ng/mL) were used as controls.
TB40/E in the presence of different concentrations of 1B2 antibody. Scale bar,
h kinetics (C) of control organoids and organoids infected with TB40/E in the
using relative organoid size. The relative organoid size for each time point is
nt mean ± SD. ***p < 0.001 by two-way ANOVA followed by Tukey’s multiple
Figure 4. NAb-Mediated Prevention of TB40/
E-Induced Abnormal Brain Organoid Struc-
ture
(A) A representative image showing layer specifi-
cation of the VZ, SVZ, and CP in brain organoids at
day 75 of differentiation.
(B) An orthogonal view of HCMV-infected brain
organoids stained for the HCMV IE1 and the
progenitor markers SOX2 and TBR2.
(C and D) Representative images of brain organo-
ids stained for the progenitor markers PAX6 (C),
SOX2 and TBR2 (D), the neuronal marker CTIP2,
and HCMV IE1. Brain organoids were mock
infected or infected with TB40/E in the presence of
IgG control antibody as organoid controls or
infected with TB40/E in the presence NAb 1B2
(2,000 ng/mL).
(E and F) Representative images of brain organoids
stained for different neuronal markers. Control
brain organoid and brain organoids infected with
TB40/E in the presence of NAb 1B2 were stained
for the cortical neuronal markers CTIP2 and BRN2
(E) and SATB2 and TBR1 (F) at day 30 post-infec-
tion.
(G) Graph comparing the relative layer thicknesses
(%) of the VZ, SVZ, andCP in control organoids and
brain organoids infected with TB40/E in the pres-
ence of 1B2 antibody. The relative thickness is
normalized to Ctrl VZ.
(H) Graph comparing the layer thicknesses of the
CTIP2-, BRN2-, SATB2-, or TBR1-positive layers in
control organoids and brain organoids treated with
TB40/E in the presence of 1B2 antibody.
Scale bars, 50 mm for (A)–(D), 100 mm for (E) and (F).
Values represent mean ± SD. **p < 0.01;
***p < 0.001 by ANOVA followed by Tukey’s multi-
ple comparison test. n = 4 organoids per group.
See also Figure S5.
Articlell
OPEN ACCESS
expression of individual receptors using receptor-specific
small interfering RNAs (siRNAs). The knockdown efficiency
of each receptor was verified in hiPSC-derived neural progen-
itor cells (NPCs) by qRT-PCR (Figure S6A). We first confirmed
the differential involvement of EGFR and PDGFRa in HCMV
entry into epithelial cells and fibroblasts, respectively (Fig-
ure S6B).42 We then treated hiPSC-derived brain organoids
at day 45 of differentiation with different receptor-specific
siRNA or a control siRNA for 4 days and subsequently
exposed the organoids to the GFP-labeled HCMV TB40/E.
Following HCMV exposure, the GFP fluorescence intensity
and the brain organoid size was evaluated for up to 20 days.
Brain organoids treated with the integrin a3 (ITGA3), a5
Cell Re
(ITGA5), or b3 (ITGB3)-specific siRNA
exhibited similar GFP fluorescence in-
tensity and growth kinetics to those of
control siRNA-treated brain organoids
(Figures 5A–5C). In contrast, brain orga-
noids treated with the EGFR- or
PDGFRa-specific siRNA showed no or
very low GFP fluorescence during the
entire 20-day post-infection period and
exhibited substantially higher growth kinetics compared to
control siRNA-treated organoids (Figures 5A–5C). These re-
sults indicate that both PDGFRa and EGFR are involved in
HCMV infection of the brain organoids. Western blot analysis
revealed that both EGFR and PDGFRa were expressed at
higher levels in NPCs than neurons (Figure 5D). To support
the importance of these receptors in HCMV infection, we eval-
uated whether overexpression of these receptors in HCMV-
resistant cells could render these cells susceptible to HCMV
infection. We prepared EGFR and PDGFRa expression vec-
tors and confirmed their expression by western blot (Fig-
ure 5E). Using these vectors, we found that overexpression
of EGFR and PDGFRa in hiPSCs, which are resistant to
ports Medicine 1, 100002, April 21, 2020 7
Figure 5. EGFR and PDGFRa Are Involved in
TB40/E Infection of Brain Organoids
hiPSC-derived brain organoids at day 45 of differ-
entiation were treated with siRNAs specific for
EGFR; PDGFRa; integrin a3, a5, or b3; or control
siRNA for 4 days and subsequently infected with
GFP-labeled TB40/E.
(A) Immunofluorescence images of TB40/E-infected
brain organoids pre-treated with different siRNAs.
Scale bar, 500 mm.
(B and C) Graphs illustrating the HCMV GFP fluo-
rescence intensity (B) and brain organoid growth
kinetics (C). The relative organoid sizes for each time
point are given as the % of the organoid size (100%)
at day 0 of infection. Values represent mean ± SD.
***p < 0.001 by two-way ANOVA followed by Tukey’s
multiple comparison test. n = 4 organoids per group.
(D) Western blot analysis of the expression of EGFR
and PDGFRa in hiPSC-derived NPCs and neurons.
GAPDH was included as a loading control.
(E) Western blot analysis showing overexpression
(OE) of EGFR and PDGFRa in HEK293T cells.
GAPDH was included as a loading control.
(F) Overexpression of EGFR and PDGFRa renders
hiPSCs susceptible to HCMV infection. hiPSCswere
electroporated with a control vector expressing RFP
or the combination of vectors expressing RFP
together with EGFR and PDGFRa and subsequently
infected with HCMV TB40/E. The HCMV-infected
cells were stained by the HCMV marker IE1. The
percent of HCMV-positive cells was quantified by
the percent of CMV IE1-positive (CMV+) cells out of
total cells. Values represent mean ± SD. ***p < 0.001
by Student’s t test. n = 4 replicates.
See also Figure S6.
Articlell
OPEN ACCESS
HCMV infection,43 rendered these cells permissive to TB40/E
infection (Figure 5F), supporting the idea that both EGFR
and PDGFRa are important mediators of HCMV infection.
These results together indicate that HCMV infection of human
8 Cell Reports Medicine 1, 100002, April 21, 2020
brain organoids involves both EGFR and
PDGFRa, whereas it does not appear to
depend on integrins, such as a3, a5, or
b3 integrin.
HCMV TB40/E Disrupts CalciumSignaling and Neural NetworkActivity in Brain OrganoidsIn order to uncover mechanisms of HCMV-
induced neurodevelopmental defects, we
performed RNA sequencing analysis
in control and HCMV TB40/E-infected brain
organoids. Interestingly, we found that at
least three out of the top ten downregulated
genes in HCMV TB40/E-infected organoids
were related tocalciumsignaling (Figure6A).
These genes include ENO2, a neuron-spe-
cific enolase that could bind to calcium;44
BNIP3, a gene involved in endoplasmic re-
ticulum (ER)/mitochondria Ca2+ homeosta-
sis;45 and PDK1, a gene involved in regula-
tion of Ca2+ entry into cells (Figure 6A).46 Gene ontology (GO)
analysis revealed that genes significantly downregulated in
TB40/E-infected brain organoids include those involved in neuro-
development, including brain development, astrocyte
Figure 6. Calcium Signaling Is Affected in HCMV-Infected Brain Organoids
RNA sequencing analysis of control and TB40/E-infected brain organoids. Brain organoids at day 45 differentiation were infected with TB40/E and analyzed by
RNA sequencing analysis 15 days post-infection.
(A) Heatmap summary of changes in mRNA expression levels in TB40/E-infected brain organoids versus UV-irradiated TB40/E-infected control brain organoids
(Ctrl).
(B) GO analysis of down- and upregulated genes in TB40/E-infected brain organoids, compared control organoids, ranked by �log10 (p value).
(C) qRT-PCR analysis of selected downregulated genes, ENO2, BNIP3, and PDK1, in TB40/E-infected brain organoids. Values represent mean ± SD. *p < 0.05,
**p < 0.01, and ***p < 0.001 by Student’s t test. n = 4 replicates.
Articlell
OPEN ACCESS
development, and hippocampal development, and genes impli-
cated in calcium-regulated exocytosis of neurotransmitter and
regulation of calcium-dependent exocytosis (Figure 6B). On the
other hand, genes significantly upregulated in TB40/E-infected
brainorganoids include those involved in immune responseand in-
Britt, University of Alabama at Birmingham) and the Vectastain ABC kit and 3,30-diaminobenzidine (DAB) substrate according to the
manufacturer’s instructions. Images were taken using a DMi8 inverted microscope equipped with a linear motorized stage.
METHOD DETAILS
Generation of brain organoids from hiPSCshiPSC-derived brain organoids were generated based on the protocol described by Lancaster et al.15 with modifications. Briefly,
hiPSCs were generated from IMR-90 and AG14048 human fibroblasts, confirmed to be karyotypically normal and negative for
mycoplasma contamination. On day 0 of organoid culture, hiPSCs were dissociated with EDTA, and seeded in suspension in a
6-well plate to form embryoid bodies in E8 medium with 5 mM ROCK inhibitor Y-27632. From day 1 to day 4, cells were cultured
in E8 medium without ROCK inhibitor with daily medium change. On day 5, E8 medium was replaced by neural induction medium
(NIM) containing DMEM-F12, 1 3 N2 supplement, 1 3 minimum essential medium NEAA (MEM-NEAA), and 2 mg/ml Heparin. On
day 8, the spheres were embedded in 20%–25%Matrigel in NIM in a 6-well suspension plate and incubated at 37�C for 4 hr, followed
by gentle addition of 2 mL of the NIM. On day 10-12, brain organoids were lifted and transferred to a new 6-well plate. NIM was
changed daily from day 5 to day 15. On day 15, brain organoids were transferred to a T25 suspension culture flask and cultured
in differentiation medium containing DMEM-F12, 1 3 N2 supplement, 2.5 mg/ml Insulin, 1 3 Glutamax, 0.5 3 MEM-NEAA, 3.5 ml/
L (V/V) 2-Mercaptoethanol, and 1 3 B27 supplement on an Orbi-Shaker (Benchmark Scientific) at 50 rpm rotating speed. Medium
was changed every 2-3 days. Organoids that exhibited similar size and passed the quality control criteria described by Lancaster
et al.15,18 were used for the study. The criteria include clear embryoid body border, formation of organized neuroepithelium before
embedding, formation of ventricle-like structure, and development of defined bud in Matrigel without premature differentiation.
HCMV infection of brain organoidsBrain organoids at day 45 of differentiation were seeded in 24-well plates and exposed to 5 3 105 pfu/ml of HCMV virus for 24 hr.
Following infection, 1 mL medium was replaced, and each organoid was placed in a single well for the duration of the experiment.
At different time points post infection, each organoid was evaluated for GFP fluorescence intensity and organoid size in diameter
using the same microscopy setting during each experiment.
Brain organoid treatment with NAbs1B2 and 62-11 NAbs at final concentrations ranging from 3.2 ng/ml to 4000 ng/ml were incubated with 5 3 105 pfu/ml of HCMV for
1 hr at 37�C and then transferred to brain organoid-containing wells. Normal mouse IgGwas used as a control at the concentration of
2000 ng/ml. For proliferation assay, brain organoidswere incubatedwith 10 mMBrdU for 2 hr and fixed in 4%paraformaldehyde (PFA)
for 1 hr, followed by immunostaining.
ImmunostainingCells on coverslip were fixed with 4% PFA in PBS for 15 min at room temperature. Brain organoids were fixed with 4% PFA for 1 hr
and submerged in 30% sucrose overnight. The samples were embedded in OCT and sectioned at a thickness of 14 or 20 mm using
Leica CM3050S. Cells or brain organoid sections were permeabilized and blocked with blocking solution (13 PBS containing 0.1%
Triton X-100 and 5%normal donkey serum) for 1 hr. Primary antibodies in blocking solution were then added and incubated at 4�C for
overnight, followed by washing and incubation with secondary antibodies. Cells were counterstained with DAPI before mounting.
The following antibodies were used: TUJ1 (Covance PRB-435P, rabbit or mouse, 1:20,000), MAP2 (Abcam ab5392, chicken,
(Accurate Chemical & Scientific Corp OBT0030, Rat, 1:6000), Cleaved Caspase3 (Cell Signaling 9661, rabbit, 1:200).
Images were obtained with a Carl Zeiss LSM700 confocal microscope or Nikon Eclipse TE-2000-S microscope. Cortical layer
thicknesses, brain organoid size and GFP fluorescence intensity of brain organoids were measured by ImageJ. For measuring the
layer thickness, small rectangles were drawn to ensure two opposite sides to align with both layer marker dashed lines.
Calcium imagingBrain organoids at day 45 of differentiation were partially dissociated and seeded on Matrigel-coated Ibidi m-slide 8-well-chamber
slides and allowed to grow for 3 days until calcium imaging was performed. Brain organoids were rinsed in artificial-cerebrospinal
fluid (ACSF) (124 mM NaCl, 2.5 mM KCl, 26 mM NaHCO3, 1 mM MgCl2, 2 mM CaCl2, 1.25 mM NaH2PO4 and 10 mM D-glucose
solution) at 37�C for 10 min and then incubated in fresh 95% O2 oxygenated ACSF containing 2 mM Fluo-4 AM for 20 min. Subse-
quently, brain organoids were visualized using a Zeiss Axio Observer Z1microscope for serial time lapse imaging. Time lapse imaging
was acquired at 10xmagnification and at 16 frames per second speed for 5min using aHamamatsu EMCCDmodel C9100-13. Gluta-
mate (3 mM) stimulation during imaging progress was performed 5 s after start. Ca2+ imaging videos were captured and processed
using ZEN software and quantification was performed using Image-Pro Premier 9.1. Fluorescence intensity change over time is
defined as: DF=F = ðF � FoÞ=Fo, where F is the fluorescence intensity at any time point, and Fo is the baseline fluorescence intensity
averaged across the wholemovie for each cell. For calcium imaging of HCMV (TB40/E-Gluc without GFP)-infected brain organoids at
Cell Reports Medicine 1, 100002, April 21, 2020 e4
Articlell
OPEN ACCESS
day 45 of differentiation, organoids were partially dissociated and allowed to attach overnight onto Matrigel-coated Ibidi m-Slide 8
Well chamber slide. Brain organoids were either exposed to HCMV at 5 3 105 pfu/ml per brain organoid or to a mixture of HCMV
and NAb 1B2. For NAb treatment, the same amount of virus was incubated with 4,000 ng/ml 1B2 antibody at 37�C for 1 hr before
addition to the brain organoid culture. As a control, brain organoids were infected with 53 105 pfu/ml UV-inactivated HCMV. Calcium
imagingwas performed 3 days post infection. In order to identify cells that were infected byHCMV, calcium imaging data of each area
were recorded with the associated coordinates using most bottom-right corner as a reference (0, 0). Right after calcium imaging, the
brain organoids were immediately fixedwith 4%PFA and immunostained for HCMV IE. For imaging, the fields were chosen based on
the calcium imaging video coordinates and the positions were adjusted manually to match the original calcium imaging video
pictures. By using this procedure, we were able to locate the HCMV positive cells by IE1 positive staining.
Microelectrode Arrays (MEA)Brain organoids at day 45 of differentiation were partially dissociated and seeded onto 12-well transparent MEA plates at three brain
organoids per well. Brain organoids were cultured in BrainPhysmedium, including 13B27, 13N2, 20 ng/ml GDNF, 20 ng/ml BDNF,
500 mg/ml Dibutyryl-cAMP, 1 3 Glutamax, and 1 3 NEAA. MEA recordings were performed at 37�C in a Maestro MEA system with
AxIS software using a bandwidth with a filter for 10Hz to 2.5 kHz cutoff frequencies. Spike detection was performed using an adaptive
threshold set to 5.5 times of the standard deviation of the estimated noise on each electrode. For recordings, following a 5min resting
time in the Maestro instrument, each plate was recorded for 10 min to calculate the spike rate per well. When a recording of 5 spikes
over the length of 1 min (5 spikes per min) was obtained, the electrode was considered active. Individual electrode bursts were iden-
tified using an adaptive Poisson surprise algorithm, while network bursts were identified for each well using a non-adaptive algorithm
requiring a minimum of 10 spikes with a maximum inter-spike interval of 100 ms. Multielectrode data analysis was performed using
the Axion Biosystems NeuralMetrics Tool. Synchrony index was calculated by NeuralMetric Tool with synchrony window set as
20ms. For the pharmacological experiment, CNQX (10 mM) or bicuculline (10 mM)were applied to plate immediately before recording.
For MEA recording, brain organoids treated with UV-irradiated TB40/E were included as the control organoids. Brain organoids were
exposed to TB40/E or UV-irradiated TB40/E at 5 3 105 pfu/ml per brain organoid in the presence of 4000 ng/ml IgG. For NAb treat-
ment, 53 105 pfu/ml of TB40/E was incubated with 4000 ng/ml NAb 1B2 for 1 hr and then added to brain organoids cultures. Wave
forms of spike were generated from exported recording data on single electrode and the graph was created in Excel. The phase
contrast and GFP fluorescent images of organoids seeded in the MEA plates were taken after MEA recording.
Generation of NPCs from human iPSCshiPSC-derived NPCs were generated according to previously described procedures75. Briefly, IMR90 hiPSCs were dissociated with
Accutase into single cells and seeded onto a Matrigel-coated 6-well plate at 1 3 105 cells per well in E8 medium containing 1 mM
Y-27632. On the next day, the E8 medium was substituted with NPC induction medium, including E6 medium, 100 nM ATRA,
10 mM SB431542, 250 nM LDN-193189, and the NPC induction medium was changed every day for 8 days. Cells were then trans-
ferred to a T25 or T75 flask and maintained in NPC maintenance medium containing 1 3 B27, 1 3 N2 supplement, 1 3 NEAA, 1 3
Glutamax, 100 nM ATRA, 3 mM CHIR99021, 2 mM SB431542, 10ng/ml EGF and 10ng/ml FGF, with daily medium change.
Neuronal differentiation from NPCsThree million hiPSC-derived NPCs were dissociated with Accutase and seeded on Matrigel-coated 10 cm plates in neural induction
medium containing DMEM/F12, 13 N2, 13 B27, 13 NEAA, 13Glutamax. Cells were cultured in this medium for 3 weeks and then
transferred to BrainPhys medium and maintained in the medium with medium change every 4-5 days.
RNA-sequencinghiPSC-derived brain organoids at day 45 differentiation were infected with 5x105 pfu/ml TB40/E or UV-irradiated TB40/E as a control.
Total RNA was isolated from TB40/E-infected organoids or control organoids 15 days post infection using Trizol. RNA quality control
was performed by the Integrative Genomics Core at City of Hope. RNA-sequencing reads were aligned against the human genome
(hg19) using TopHat276. Read counts were quantified using htseq-count with UCSC known gene annotations. Aligned reads were
counted using GenomicRanges. Genes were filtered to only include transcripts with RPKM values greater than 0.1 (after a rounded
log2-transformation) in at least 50% samples. Genes smaller than 150 bpwere removed prior to differential expression analysis. Log2(RPKM + 0.1) expression values were used for visualization and fold-change calculations. Normalization of heatmap values was
performed as the following: RNA seq of control sample was converted by Log10 of the RPKM value, the HCMV infected sample value
was determined by sum= log10ðRPKM valueÞ+DEseq2� log2ðfold change valueÞ. Heatmap was generated by ClustVis and image
was prepared using Photoshop.
RT-qPCR analysisFor siRNA knockdown of receptors in iPSC-derived NPCs, total RNA was extracted using Trizol Reagent. cDNAs were reverse
transcribed using Tetro cDNA synthesis kit. RT-qPCRwas performed usingDyNAmo Flash SYBRGreen qPCRmix on a StepOnePlus
system and normalized to b-actin. Primers used in the qPCR are listed in Key Resources Table.
e5 Cell Reports Medicine 1, 100002, April 21, 2020
Articlell
OPEN ACCESS
RNA interferenceOligonucleotides for siRNA-mediated RNA interference were synthesized by Integrated Device Technology. Oligonucleotides used
for RNA interference were listed in the Key Resources Table. For RNA interference in brain organoids, four brain organoids with
similar size were seeded in ultralow 24-well plates (one organoid per well) and transfected with siRNA using siLentFect. After
96 hr incubation, each organoid was infectedwith TB40/E at 53 105 pfu/ml. Mediumwas changed the following day. Brain organoids
were cultured for 20 days on shaker with continuous medium change every 3-4 days.
Western blot analysisCells were lysed in 0.1 M Na2CO3 containing 2 mM PMSF, and cell lysates were sonicated using Sonic Dismembrator to disintegrate
genomic DNA. Protein concentrations were measured by Bradford Assay. Forty mg total proteins of NPCs and neurons were loaded
for western blot analysis. EGFR and PDGFRa antibodies were used at 1:1000 dilution.
PlasmidsEGFR WT was a gift from Matthew Meyerson (Addgene plasmid # 11011; http://addgene.org/11011; RRID:Addgene_11011). The
pHIV7-EGFR-T2A-DsRed vector was generated by cloning the human EGFR coding region via SwaI and NotI sites into the plasmid
pHIV-T2A-DsRed (kindly provided by Dr. Xiuli Wang from City of Hope), resulting in the EGFR expression vector. pDONR223-
PDGFRA was a gift fromWilliam Hahn & David Root (Addgene plasmid # 23892; http://addgene.org/23892; RRID:Addgene_23892).
The pcDNA-PDGFRa vector was created by cloning the human PDGFRa cDNA into pcDNA3.1(+) neo vector using NheI and SmaI
sites. The sequences of the cDNA were verified by Sanger sequencing.
hiPSC electroporation and HCMV infectionControl DsRed plasmid DNA or the mix of EGFR-T2A-DsRed and PDGFRa plasmid DNAs were introduced into hiPSCs by electro-
poration using 4D-Nucleofector following manufacturer’s instructions. Briefly, 3 million hiPSCs in single cell suspension were mixed
with P3 solution and subjected to electroporation using program CA137. The electroporated iPSCs were immediately transferred
onto Matrigel-coated 24-well plate in E8 medium containing 1mM Y27632. The hiPSCs were subcultured twice in order to remove
dead cells and subsequently allowed to attach on a Matrigel-coated 24-well plate at 2x104 cells/well. hiPSCs were infected with
TB40/E at a MOI of 5 and harvested at 2 days after infection. The percent of CMV IE1-positive and DsRed-positive (CMV+DsRed+)
cells out of total DsRed+ cells was quantified and plotted.
QUANTIFICATION AND STATISTICAL ANALYSIS
Statistical analysisAll statistical analyses were performed using GraphPad Prism7.0. Statistical details of experiments can be found in the figure leg-
ends. All data are shown as mean ± SD or SE. The statistical significance of experiment outcome when comparing two or more
groups was calculated using two-way ANOVA followed by Tukey’s, Dunnet’s or Sidak’s multiple comparison tests or Student’s
t test. The difference between experimental groups was considered significant when p < 0.05.
DATA AND CODE AVAILABILITY
The RNA-seq dataset generated during this study is available at NCBI. The GEO Accession Super Series ID is GEO: GSE145415.
Cell Reports Medicine 1, 100002, April 21, 2020 e6
Guoqiang Sun, Flavia Chiuppesi, Xianwei Chen, Cheng Wang, E Tian, JennyNguyen, Mindy Kha, Daniel Trinh, Hannah Zhang, Maria C. Marchetto, HongjunSong, Guo-Li Ming, Fred H. Gage, Don J. Diamond, Felix Wussow, and Yanhong Shi
1
Supplemental Figures and Legends
Figure S1. Functional neuronal networks shown by MEA recording. Related to Fig. 1. Quantification of MEA
parameters in brain organoids in 10 min recording before or after treatment with GABAergic neuronal inhibitor
bicuculine (Bicu), and before or after glutamatergic neuronal inhibitor CNQX treatments.
2
Figure S2. HCMV cell tropism and HCMV-induced abnormal cell proliferation and apoptosis in hiPSC-
derived brain organoids. Related to Fig. 2. (A) ARPE-19 epithelial cells and MRC-5 fibroblasts were seeded in
96-well plates and either mock-infected or infected with TB40/E or Towne. HCMV-infected cells were
immunostained for HCMV IE1 at 24 hours post infection. Shown are representative images of the IE1
immunostaining. Scale bar: 500 µm. (B-E) Quantification of the number of proliferating and apoptotic cells in brain
organoids infected with the GFP-tagged HCMV TB40/E (B and C) or the non-GFP tagged HCMV TB40/E-Gluc (D
and E) in comparison to mock-infected brain organoid controls. The number of proliferating cells was evaluated by
the percent of BrdU+ cells out of total cells, while apoptosis was assessed by the percent of active caspase 3-positive
(Active Cas3+) cells out of total cells. Bars represent mean ± SD. **p<0.01 and ***p<0.001 by Student’s t-test .
n=4 replicates.
3
Figure S3. Prevention of HCMV TB40/E-induced abnormal brain organoid growth by Nab 62-11. Related to
Fig. 3. hiPSC-derived brain organoids at day 45 of differentiation were infected with GFP-labeled TB40/E in the
presence of different concentrations of NAb 62-11 that ranged from 3.2 ng/ml to 2000 ng/ml of antibody. Mock-
infected brain organoids and TB40/E-infected organoids treated with IgG control (2000 ng/ml) were used as control
organoids. (A) Representative images of control organoids and brain organoids infected with TB40/E in the
presence of different concentrations of NAb 62-11. The images of the control organoids are the same as that in Fig.
3 because the experiments in Fig. 3 and Fig. S3 were performed in parallel. Scale bar, 200 µm. (B, C) Graphs
illustrating the HCMV-GFP fluorescence intensity (B) and growth kinetics (C) of control organoids and TB40/E-
infected brain organoids treated with NAb 62-11. Growth kinetics is measured using relative organoid size. The
relative organoid size for each time point is given as the % of the organoid size (100%) at day 0 of infection. Values
represent mean ± SD. *** p<0.001 by two way ANOVA followed by Tukey’s multiple comparison test. n=4
organoids per group.
4
Figure S4. NAb prevents abnormal brain organoid growth induced by TB40/E and TR in hiPSC-derived
organoids. Related to Fig. 3. Human IMR90 iPSC-derived brain organoids at day 45 of differentiation were
infected with GFP-labeled TB40/E or TR in the absence or presence of NAb. Mock-infected brain organoids were
used as a control. (A) Representative images of brain organoids that were mock-infected or infected with TB40/E in
the presence of IgG (Mock + IgG, or TB40/E + IgG) or brain organoids infected with TB40/E in the presence of
NAb 1B2 (TB40/E + 1B2) at the given time points during 30 days post infection. Scale bar, 200µm. (B, C) Graphs
indicating the relative GFP fluorescence intensity (arbitrary units) (B) or relative organoid size (C) in mock-infected
control brain organoids, or brain organoids infected with TB40/E in the absence or presence of NAb 1B2. (D, E)
Graphs indicating the relative GFP fluorescence intensity (arbitrary units) (D) or relative organoid size (E) in mock-
infected control brain organoids, or brain organoids infected with TR in the absence or presence of NAb 1B2. The
relative organoid size for each time point is given as the % of the organoid size (100%) at day 0 of infection. Values
represent mean ± SD. *** p<0.001 by two way ANOVA followed by Tukey’s multiple comparison test. n=4
replicates.
5
Figure S5. NAb prevents TB40/E-induced abnormal layer composition in brain organoids. Related to Fig. 4.
(A, B) Representative images (cropped from images in Fig. 4C, D) were used for counting specific layer marker-
positive cells. Scale bar, 20 µm. (C) Quantification of PAX6+, TBR2+, and CTIP2+ cells in brain organoids that
were mock-infected in the presence of IgG control or infected with TB40/E in the presence of IgG control or NAb
1B2, in images shown in panels A and B. p > 0.05 (ns), **p < 0.01 by one-way ANOVA followed by Tukey’s
multiple comparison test. n=4 replicates. (D) Quantification of PAX6+, TBR2+, and CTIP2+ cells in mock-infected
control organoids and brain organoids infected with non-GFP-tagged TB40/E-Gluc or TR. p > 0.05 (ns), **p < 0.01,
***p < 0.001 by Student’s t-test. n=4 replicates. For panels C & D, values represent mean ± SD.
6
Figure S6. Knockdown of receptor gene expression by receptor-specific siRNAs. Related to Fig. 5. (A)
Evaluation of siRNA efficacy. hiPSC-derived NPCs were transfected with siRNAs specific for PDGFRα, EGFR,
ITGA5, ITGB3, ITGA3, or non-targeting control siRNA (siC). At four days post transfection, mRNA expression of
specific genes was analyzed by RT-qPCR. Values represent mean ± SD. ***p < 0.001 by Student’s t-test. n=4
replicates. (B) siRNA-mediated inhibition of HCMV infection of fibroblast and epithelial cells. ARPE-19 and MRC-
5 cells were treated with the receptor-specific siRNAs or siC. Following 48-hour incubation, cells were infected
with TB40/E or Towne and stained for HCMV IE1 at 24 hours post-infection. The percent of HCMV-infected cells
was calculated relative to the number of IE1-positive cells in HCMV-infected cells treated with siC. Values
represent mean ± SEM of three independent experiments performed in triplicate wells. Statistical significance was
calculated to HCMV/siCTRL group using 2-way ANOVA followed by Dunnett's multiple comparisons test.
*p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. n=3 replicates.
7
Figure S7. Representative images of brain organoids seeded on MEA electrodes. Related to Fig. 7. Sample
images of mock-infected brain organoids and brain organoids infected with GFP-tagged TB40/E in the absence or